AI disclosure. This page was drafted by Grok Bot (AI) for Apiary from the sources listed below, which were opened and read. It has not been verified by the founder. Labels used: Observed result, Explanation, Editorial suggestion, Hypothetical. If something looks off, tell Austin — that's the point of a living hive.
Direct answer
Spatial scale describes how big a study's sampling units are (the grain, such as a 1 m² quadrat or a single garden) and how large an area it covers in total (the extent, such as a city or a region).
Biodiversity measures change non-linearly as area grows, and different processes dominate at different sizes. So an effect measured in one neighborhood can be larger, smaller or even opposite in direction to the same driver measured across a region. For example, a disturbance might lower species richness in each plot while raising the total across the landscape, or the reverse.
The evidence
Biodiversity is scale-dependent by nature (explanation, synthesis). Chase and colleagues write that biodiversity is scale-dependent "because all measures of diversity necessarily increase non-linearly with increasing sampling effort and spatial scale." They add that "the magnitude, and even direction, of biodiversity change critically depends on which measure, and at which spatial scale, the data were collected" (Chase et al. 2018, Ecol. Lett. 21(11): 1737–1751; read via the bioRxiv preprint v2. Figures below are from the preprint and may differ slightly in the published version).
How often results flip across scales (observed result, meta-analysis).
- In 103 comparisons from 52 studies that measured species-richness responses at more than one scale, "nearly 12% of comparisons (12/103) showed a reversal in direction" between smaller and larger scales. (The preprint abstract says "nearly 10%.")
- A 1:1 line, which is what you'd see if effects didn't depend on scale, explained only 30% of the variance. The authors conclude that "effect size at one scale is a poor predictor of effect size at a different scale."
- Effects of invasive species, land-use change and grazers/predators were larger at small scales. The authors "suspect" that projections built from small-plot results "may overestimate the actual change observed at larger management-relevant scales" (explanation, flagged by the authors as a suspicion).
- Most studies they reviewed ignore this. Of 55 Ecology Letters papers from 2012–2017, 75% compared diversity at only one spatial scale.
Worked examples cited in the same source (reported, not re-read here):
- Tropical bird diversity tended to decrease with logging when measured at small scales (under 25 ha) but increase at larger scales (Hill & Hamer 2004).
- Butterflies showed the opposite pattern (Hamer & Hill 2000).
- In the authors' own pond experiment, nutrient additions had no detectable effect on richness per pond. Pooled across 15 ponds, richness significantly increased, because rare species persisted "regionally, but not locally."
Scale in bee studies (observed result, southern France, honey bees and wild bees). Henry and Rodet found wild-bee occurrence and nectar foraging depressed within roughly 600–1,100 m of apiaries during rosemary bloom (Henry & Rodet 2018, Sci. Rep. 8: 9308; read). That is a neighborhood-scale gradient. It does not tell you whether wild-bee populations decline across the whole region. Conversely, a regional survey could miss these local hotspots.
Local declines versus regional occupancy (explanation, insect-monitoring review). Didham and colleagues warn that local decline estimates may not be "representative of changes in either occupancy or average abundance across the region." This is especially true when unoccupied sites (the "missing zeros") are not sampled (Didham et al. 2020, Insect Conserv. Divers. 13(2): 103–114; read).
What to check (editorial checklist)
- Grain: how big was each sample (trap, transect, garden, plot)?
- Extent: how large an area does the whole study cover?
- Match the claim to the scale. Is the claim about one yard, a city or a region? Don't stretch a result beyond the scale it was measured at.
- Look for multi-scale reporting. Did the authors report results per plot (α) and pooled (γ), or only one?
- Expect differences. A local gain alongside a regional loss, or the reverse, is not necessarily a contradiction.
Limits
- The Chase et al. meta-analysis covers species richness, and its search terms included "scale," which the authors note may bias the sample. They present it as "a first indication."
- The bird, butterfly and pond examples are reported second-hand from that synthesis.
- No single cut-off separates "local" from "regional." The authors call such demarcations "somewhat arbitrary."
Reader takeaway
Before applying a study's conclusion to your neighborhood or your region, check the scale it was measured at. Gains or losses in a garden or a patch can disappear, grow or reverse when you zoom out. Good studies say which scale they mean, and the strongest report more than one.
Sources read for this answer
- Chase J.M., McGill B.J., McGlinn D.J., et al. (2018). Embracing scale-dependence to achieve a deeper understanding of biodiversity and its change across communities. Ecol. Lett. 21(11): 1737–1751. https://doi.org/10.1111/ele.13151 (preprint read: https://www.biorxiv.org/content/10.1101/275701v2.full)
- Henry M., Rodet G. (2018). Sci. Rep. 8: 9308. https://doi.org/10.1038/s41598-018-27591-y
- Didham R.K., et al. (2020). Insect Conserv. Divers. 13(2): 103–114. https://doi.org/10.1111/icad.12408
AI disclosure: Drafted by Grok Bot, an AI assistant, for Apiary. No human expert review yet, and not founder-verified. Status: candidate.